IBM 0X9B0D RAID Controller Compatible Battery 7.4V 6800mAh

IBM 0X9B0D / 0XC9F3 RAID Controller Cache Battery — 7.4V Li-ion Replacement

This 7.4V Li-ion cell at 6800mAh (50.32Wh) replaces the backup cache battery in IBM RAID controllers including the 0X9B0D, 0XC9F3, 271-00011, and AVT-900486 series. It powers the controller's volatile write cache during a power loss, holding data long enough for a safe flush to disk. Without a functional cell, most controllers drop to write-through mode and cache acceleration stops.

• Cross-model fitment — 0X9B0D, 0XC9F3, 271-00011, AVT-900486: These controllers share the same 7.4V battery rail, connector pinout, and BMS handshake protocol, which is why one cell covers the full group. The controller polls cell voltage and chemistry flags at boot — mismatches on either cause an immediate fault.
• Bench tested on actual hardware: We ran charge and discharge cycles through the BMS on this cell and confirmed the protection thresholds fire correctly. The BMS reported cell status cleanly over the SMBus interface with no fault flags during testing.
• Post-swap controller calibration: After installing the new cell, trigger a battery learn cycle from the management interface — on ONTAP use system controller battery show, on Broadcom/LSI use StorCLI /cx bbu show. The controller holds write-through mode until the learn cycle finishes and recalculates the backup window against the new cell's actual capacity.

RAID controller staying in write-through mode after battery swap

Swapping the cell does not automatically restore write-back mode. The controller requires a completed learn cycle before it trusts the new battery enough to re-enable the write cache. This cycle runs a controlled full charge and discharge to measure the actual backup window the new cell can deliver. Until the cycle completes — typically 24 to 72 hours under normal system load — the controller stays conservative and keeps write-through active.

Management console reporting a battery error days after a clean installation

This happens because the controller runs its battery health assessment on a timed schedule, not at boot. The first scheduled self-test after installation may flag the battery if the learn cycle has not yet completed and the backup window estimate is still uncalibrated. The controller is not reporting a faulty cell — it is reporting an unverified one. Let the learn cycle run to completion, then check status again with StorCLI /cx bbu show or the equivalent ONTAP command to confirm the error clears.